five

Trelis/eval-ursa-2-enhanced-multimed-hard-20260408-1933

收藏
Hugging Face2026-04-08 更新2026-04-12 收录
下载链接:
https://hf-mirror.com/datasets/Trelis/eval-ursa-2-enhanced-multimed-hard-20260408-1933
下载链接
链接失效反馈
官方服务:
资源简介:
--- tags: - whisper - evaluation - speech - speech-to-text --- # Evaluation Results: ursa-2-enhanced Evaluation results from Whisper model evaluation. ## Summary | Model | WER | CER | |-------|-----|-----| | [speechmatics/ursa-2-enhanced](https://huggingface.co/speechmatics/ursa-2-enhanced) | 10.46% | 6.03% | ## Source Data - **Evaluation Dataset:** [Trelis/multimed-hard](https://huggingface.co/datasets/Trelis/multimed-hard) - **Model Evaluated:** [speechmatics/ursa-2-enhanced](https://huggingface.co/speechmatics/ursa-2-enhanced) ## Columns | Column | Description | |--------|-------------| | `audio` | Audio sample (if available from source dataset) | | `reference` | Ground truth transcription | | `prediction` | Model prediction | | `wer` | Word Error Rate for this sample | | `cer` | Character Error Rate for this sample | | `entities` | Entity annotations from source dataset | | `entity_cer` | Per-sample entity CER (-1.0 if no entities) | ## Entity CER **Overall Entity CER:** 19.55% | Category | CER | |----------|-----| | anatomy | 25.11% | | biomarker | 15.13% | | condition | 14.82% | | drug | 0.00% | | organisation | 44.55% | | procedure | 20.71% | --- *Generated by [Trelis Studio](https://studio.trelis.com)*

--- tags: - Whisper语音识别模型(Whisper) - 评测 - 语音 - 语音转文字(Speech-to-Text) --- # 评测结果:ursa-2-enhanced 本结果来自Whisper模型的评测。 ## 总结 | 模型 | 词错误率(Word Error Rate, WER) | 字符错误率(Character Error Rate, CER) | |-------|-----|-----| | [speechmatics/ursa-2-enhanced](https://huggingface.co/speechmatics/ursa-2-enhanced) | 10.46% | 6.03% | ## 源数据集 - **评测数据集:** [Trelis/multimed-hard](https://huggingface.co/datasets/Trelis/multimed-hard) - **被评测模型:** [speechmatics/ursa-2-enhanced](https://huggingface.co/speechmatics/ursa-2-enhanced) ## 字段说明 | 字段名 | 描述 | |--------|-------------| | `audio` | 音频样本(若源数据集提供) | | `reference` | 基准真值转录文本 | | `prediction` | 模型预测转录文本 | | `wer` | 单样本词错误率 | | `cer` | 单样本字符错误率 | | `entities` | 源数据集自带的实体标注 | | `entity_cer` | 单样本实体字符错误率(若无实体则取值为-1.0) | ## 实体字符错误率 **整体实体字符错误率:** 19.55% | 类别 | CER | |----------|-----| | 解剖学(anatomy) | 25.11% | | 生物标志物(biomarker) | 15.13% | | 病症(condition) | 14.82% | | 药物(drug) | 0.00% | | 组织机构(organisation) | 44.55% | | 诊疗操作(procedure) | 20.71% | --- *由 [Trelis Studio](https://studio.trelis.com) 生成*
提供机构:
Trelis
5,000+
优质数据集
54 个
任务类型
进入经典数据集
二维码
社区交流群

面向社区/商业的数据集话题

二维码
科研交流群

面向高校/科研机构的开源数据集话题

数据驱动未来

携手共赢发展

商业合作